RECENT ADVANCES IN SCIENCE 523 



the fundamental conceptions and the hitherto accepted laws 

 of dynamics. The foundations of classical mechanics, as enun- 

 ciated in Newton's Laws of Motion, are laid in an unyielding 

 soil of Euclidean space of the type we are made familiar with 

 from the first beginnings of our geometrical experience and 

 teaching, using what is now termed a Galilean frame of refer- 

 ence. Einstein's philosophical and mathematical reasoning, 

 and the results he obtained, verified d priori like that referring 

 to the perihelion of Mercury, or d posteriori like the bending 

 of rays of light when passing through a strong gravitational 

 field, can be treated in one of two ways. The counsel of 

 caution is to accept the new facts thus established and to 

 incorporate them in the traditional system of physical science. 

 Thus, the bending of light is to be taken as establishing a 

 connecting link between light and gravitation, whilst the ex- 

 planation of the motion of the perihelion of Mercury is to be 

 taken as showing that Newton's law of gravitation needs 

 modification when great accuracy is aimed at. The bolder 

 view is to accept the mighty structure of Einstein's theory, 

 with the embarrassing consequence that space loses its invari- 

 able Euclidean character — in fact, Euclidean space no longer 

 exists anywhere, since every part of what we call space is 

 occupied by a gravitational field ; there is, further, the additional 

 consequence that the notion of force as underlying motion loses 

 all meaning, what we have hitherto called force being merely 

 a reflex of the properties of the space in which the motion 

 takes place. 



No doubt it seems easier for the applied mathematician to 

 adopt the counsel of caution, for this is compatible with the 

 survival of the classical notions of mass and force. It is, 

 however, hardly a logical proceeding to accept new facts 

 without working these facts into a system consistent with 

 our former possessions. The old Newtonian view of gravita- 

 tion possessed the charm of simplicity, and the advantage 

 of fitting in with the Euclidean space that underlies classical 

 mechanics. Moreover, the bending of light in a gravitational 

 field presents the difficulty that the predicted and verified 

 amount of bending is double what we would be led to expect 

 if we merely endow visual radiation withgravitational properties. 



If there are difficulties about adopting the view that the 

 bending of light and the explanation of the motion of Mercury 

 merely add new facts to the established system of mechanics, 

 a revolution in our ideas is involved in the acceptance of the 

 complete superstructure of theory that has been raised on these 

 facts. We are concerned here with the mechanical conse- 

 quences of the theory, and, put in brief, they appear to be as 

 follows. 



